Biomimetic sequentially gated nanotuners based on amorphous metal–organic frameworks for reprogramming the metabolism-ferroptosis-immunity crosstalk in gliomas

TL;DR

A new nanotuner (ZB@HM) reprograms glioma metabolism and immunity to enhance ferroptosis and antitumor immunity.

Contribution

A biomimetic sequentially gated nanotuner that modulates metabolism-ferroptosis-immunity crosstalk in gliomas.

Findings

ZB@HM triggers ferroptosis and immunogenic cell death via NADPH and cysteine depletion.

Methionine restriction by ZB@HM suppresses tumor cysteine biosynthesis and enhances T-cell methionine uptake.

Treatment with ZB@HM downregulates immune checkpoints and reverses T-cell exhaustion.

Abstract

Glioma's metabolic reprogramming fortifies antioxidant defenses and fosters an immunosuppressive microenvironment, thus resulting in robust resistance to ferroptosis-immunotherapy. In this study, a biomimetic sequentially gated amorphous MOF-based nanotuner (ZB@HM) was constructed, which could cascade depletion of metabolic substrates of ferroptosis defense and redistribution of methionine for reprogramming metabolism-ferroptosis-immunity crosstalk. Specifically, we engineered a methionine-inhibitor-loaded amorphous ZIF-82, followed by coating with E. coli-glioma hybrid membrane to improve its blood-brain barrier penetration and gliomas-targeting. The amorphous MOF undergoes acidic-triggered disintegration, releasing 2-nitroimidazole that is selectively activated via NADPH reduction in hypoxia. The activated ligand subsequently covalently binds with thiol-containing cysteine. The cascade depletion of NADPH and cysteine, synergizing with Zn2+ overload, effectively triggers robust ferroptosis and immunogenic cell death. Moreover, ZB@HM selectively restricts methionine uptake in gliomas, suppressing the transsulfuration pathway for cysteine biosynthesis and enhancing the competitive uptake of methionine by T cells. The reduced intracellular methionine pool in gliomas diminishes S-adenosylmethionine biosynthesis, downregulating immune checkpoint expression. This cascade reverses T-cell exhaustion and reinforces antitumor immunity. RNA sequencing analysis revealed that treatment with ZB@HM resulted in a modulation of gene signatures associated with the GSH metabolism and tumor immunotherapy. Collectively, ZB@HM reprograms the metabolism-ferroptosis-immune crosstalk through orchestrating multiple key metabolites.

This study synthesizes a biomimetic sequentially gated (acidity-hypoxia responsive) MOF-based nanotuner (ZB@HM) that remodels immunosuppressive microenvironment by reprogramming metabolism-ferroptosis-immunity crosstalk network. Mechanistically, ZB@HM undergoes acid-mediated degradation, releasing 2-nitroimidazole, which disrupts ferroptosis defense via hypoxia-responsive NADPH/cysteine cascade depletion. Released methionine uptake inhibitors selectively restricts tumor's methionine supply, potentiating ferroptosis, downregulating immune checkpoints and reversing T-cell exhaustion.Image 1

Figures

a) TEM image of ZIF-82, scale bar: 500 nm; TEM images of b) ZB and c) ZB@HM, scale bar: 100 nm; d) SDS-PAGE analysis for membrane proteins; e) HAADF-STEM image and elemental mapping of Zn in ZB@HM; f) N_2_ adsorption-desorption isotherms of ZB; g) Zn 2p spectrum of ZB; h) Zeta potential and i) hydrodynamic diameters of ZB and ZB@HM. Data are mean ± SD; j) TEM images of ZB@HM in a pH 6.5 buffer solution at 0 min, 30 min, 1 h, and 8 h; The time-dependent release percentage of k) BCH, l) Zn^2+^ and m) 2-nim from ZB@HM at different pH values (pH 7.4, pH 6.5, pH 5.0); n) Hydrodynamic diameters of Z

a) Schematic diagram of *in vitro* BBB model for assessing the capacity of NPs to across the BBB; b) Fluorescent microscopy images showing localization of FITC-labeled NPs in GL261 cells following transport across an *in vitro* BBB model. Scale bar: 50 μm; c) Fluorescent microscopy images showing cellular uptake of Rhodamine B-labeled NPs by GL261 cells assessed at the designated time points. Scale bar: 50 μm; d) Viability of BV2 and GL261 cells after exposure to NPs of different concentrations in environments with distinct pH values; e) Colony formation assay images of GL261 cells across dist

a) Fluorescent microscopy images of GL261 cells in different groups stained with Zn^2+^ indicator probe, coupled with the quantification of the probe's fluorescence intensity (1: Control, 2: BCH, 3: ZIF-82, 4: ZB, 5: ZB@HM). Scale bar: 50 μm; b) Fluorescent microscopy images of GL261 cells in different groups stained with JC-1, and the quantification of the ratio between JC-1 aggregates and monomers. Scale bar: 50 μm; c) Fluorescent microscopy images of GL261 cells in different groups stained with DCFH-DA, and quantification of the fluorescence intensity of DCFH-DA. Scale bar: 50 μm. Bar graph

Fluorescent microscopy images of GL261 cells in different groups stained with a) HMGB1, c) CRT. Scale bar: 50 μm; Bar graph depicting the mean fluorescence intensity (MFI) of b) HMGB1, and d) CRT in different treatment groups (1: Control, 2: BCH, 3: ZIF-82, 4: ZB, 5: ZB@HM); e) The relative extracellular ATP (eATP) content in GL261 cells for different treatment groups; f) The relative SAM content in GL261 cells for different treatment groups; g) Western blot analysis of proteins *in vitro* related to immune, including PD-L1, GAPDH, H3K4Me3, and Histone H3; h) Representative flow cytometry anal

a) The volcano map. b) Gene Ontology enrichment analysis about representative pathways among the identified gene set. c) KEGG analysis about representative pathways among the identified gene set. d) GSH metabolism. e) The PD-1 and PD-L1 checkpoint pathway in tumor.Fig. 5

a) Timeline illustration depicting the experimental procedure for glioma-related research in mice; b) MRI scans of mice brains in different treatment groups at various time points; c) Representative micrographs of brain tissue sections derived from distinct treatment groups. scale bar = 1000 μm; d) High-magnification histological micrographs of brain tissues derived from distinct treatment groups. scale bar = 100 μm; e) Relative tumor volume curves of brain tumor from different treatment groups. f) Western blot analysis of proteins *in vivo* related to ferroptosis, including SLC3A2 (4F2hc), SL

a) Relationship of SLC7A5 expression and immune cell infiltration levels based on information extracted from the Tumor Immune Estimation Resource (TIMER) database. b) Representative assessment of CD86 and CD80 expression in mouse tumor tissues via flow cytometry with different groups, and bar graph showing the percentage of DC cells for different treatment groups (1: Control, 2: BCH, 3: ZIF-82, 4: ZB, 5: ZB@HM); c) Representative Flow cytometric analysis of CD3 and CD4 expression in mouse tumor tissue, and bar graph showing the percentage of CD4^+^ T cells (1: Control, 2: BCH, 3: ZIF-82, 4: ZB

Full text

Introduction

1

Gliomas are the most common and aggressive central nervous system tumors, and its efficient therapy remains a huge challenge. In recent years, cancer immunotherapy centered on immunogenic cell death (ICD) has attracted increasing attention as a highly promising therapeutic strategy [1]. Ferroptosis, a regulated form of cell death with the capacity to induce ICD, has been identified as a prospective therapeutic target in gliomas by exploiting the high dependence of gliomas on lipid metabolism [[2], [3], [4]]. However, metabolic reprogramming of glioma drives robust cellular defense mechanisms and highly tumor immunosuppressive microenvironment (TIME), leading to suboptimal therapeutic efficacy of ferroptosis-immunotherapy [5,6]. Addressing these challenges requires the development of innovative strategies to precisely disrupt key metabolites in ferroptosis defense and remodel the TIME.

The sustained supply of diverse reductive substrates—notably nicotinamide adenine dinucleotide phosphate (NADPH) and cysteine—represents a key mechanism underlying the establishment of robust ferroptosis defense systems in tumor cells. NADPH is the core of the ferroptosis defense axis by providing the reducing power necessary for cellular antioxidant systems [7]. Cellular availability of cysteine, as the limiting precursor of GSH synthesis, is a primary determinant of the activity of glutathione/glutathione peroxidase 4 (GPX4) axis that defends against ferroptosis [8,9]. Thus, it is highly desired to disrupt the metabolic equilibrium by efficiently depleting both NADPH and cysteine. Inspired by the fact that strong electron-withdrawing nitroimidazole could mimic the "oxygen effect" to oxidize NADPH and abstract hydrogen atoms from the sulfhydryl groups of cysteine [10,11], it is a promising strategy of amplifying ferroptosis to developed a metal-organic framework (MOF) with nitroimidazole-derived ligands and ferroptosis-inducing metal ions for cascaded depleting NADPH and cysteine, while simultaneously establishing self-amplifying positive feedback loops.

An additional critical determinant governing the efficacy of ferroptosis-induced immune activation is the immunophenotypic landscape of the glioma microenvironment. The methionine cycle functions as a pivotal metabolic hub in oncology, coordinating multifaceted oncogenic processes that encompass transcriptional upregulation of immune checkpoint proteins, induction of T cell dysfunction, and collective promotion of immune evasion mechanisms [[12], [13], [14]]. However, non-specific depletion of methionine fails to sustain normal methionine uptake in cytotoxic T lymphocytes situated within the tumor microenvironment (TME), and this impairment subsequently affects the effector functions of T cells [15]. Thus, a critical challenge is to selectively constrain methionine uptake in tumor tissues, while simultaneously boosting the competitive acquisition of methionine by T cells. Notably, methionine serves as an essential substrate for the transsulfuration pathway that generates cysteine. Its targeted depletion in tumors is therefore anticipated to suppress GSH biosynthesis, thereby providing a complementary mechanism to amplify ferroptosis [16].

The blood-brain barrier (BBB) critically restricts drugs to penetrate into the central nervous system, limiting their therapeutic efficacy [17]. In recent years, the biomimetic strategies utilizing cell membrane-camouflaged nanoparticles to inherit the biological functions have been widely employed in targeted drug delivery [[18], [19], [20]]. Binding of outer membrane protein A (OmpA)—a component of the gram-negative bacterium Escherichia coli K1 (EC-K1)—to surface-expressed gp96 on BBB endothelial cells mediates the invasion of these endothelial cells by EC-K1 [21,22]. Tumor cell membranes with “homing effect” endow the nanomaterials with good ability of target and recognizing the tumors in vivo [23]. Therefore, a biomimetic approach based on the bacterial–tumor hybrid membrane is great potential for improving the BBB penetration and targeting ability of nanomaterials. However, no single strategy has been developed to deplete metabolic substrates of ferroptosis defense and redistribution of methionine, thus reprogramming metabolism-ferroptosis-immunity crosstalk.

Herein, we fabricated a biomimetic nanotuner (ZB@HM) based on a sequentially gated (acidic and hypoxic TME) amorphous MOF (ZIF-82) loading with the methionine uptake inhibitors (BCH), followed by cloaking with E. coli/glioma hybrid cell membrane (Scheme 1). E. coli-derived OmpA mediates BBB transcytosis, while glioma membrane components confer tumor homing, enabling ZB@HM to pass through BBB and target glioma [24,25]. Under the acidic TME, ZB@HM degrades and releases electron-withdrawing ligand 2-nitroimidazole (2-nim), Zn^2+^ and BCH [26]. Under hypoxia, 2-nim could be reduced to 2-hydroxylaminoimidazole by NADPH, and further forms stable adducts with thiol (–SH) groups in cysteine, cascade depleting the two pivotal ferroptosis substrates–NADPH and cysteine [27,28]. Combined with ROS produced by Zn^2+^ overload drives the robust ferroptosis and activates ICD. In addition, ZB@HM mediates tumor-selective methionine uptake blockade, which not only depletes cysteine reservoirs to further potentiate ferroptosis, but also concurrently reduces intracellular S-adenosylmethionine (SAM) pools in malignant cells, thereby eliciting transcriptional downregulation of immune checkpoint molecules through epigenetic remodeling. It concomitantly reprograms intercellular methionine partitioning to preferentially sustain T cell metabolic demand, revitalize exhausted T cells, and ultimately potentiate systemic antitumor immunity. RNA sequencing analysis identified ZB@HM-induced transcriptional reprogramming of GSH metabolism-related genes and immunotherapy-relevant signatures. More importantly, we present an old drug repurposing strategy for extending the arsenal of oncology therapies. The ZB@HM is biodegradable and can be readily synthesized, thereby facilitating its future application development. Overall, these findings demonstrate that ZB@HM could effectively cascade deplete dual-substrate to disrupt ferroptosis defenses, boost the immune response and impede glioma growth by leveraging the synergistic interaction of ferroptosis and immunotherapy.Scheme 1. The schematic illustration of the a) development of ZB@HM, and therapeutic mechanism of ZB@HM by b) Cascade depletion of both NADPH and cysteine; c) Robust ferroptosis-triggered ICD; d) Methionine redistribution that enhances ferroptosis, suppress PD-L1 and restore T-cell viability and e) alleviation of tumor immunosuppression.Scheme 1

Experimental methods

2

Materials

2.1

Zinc nitrate hexahydrate, 2-Nitroimidazole (2-nIm, 98%), 1H-Imidazole-4-carbonitrile (cnIm, 98%), PVP58000, and N, N-Dimethylformamide (DMF) were acquired from Aladdin biochemical Technology Co., Ltd. (Shanghai, PR China). Both Cell Counting Kit-8 (CCK-8) and fluorescein isothiocyanate isomer (FITC-I) were sourced from Beijing Bioss Biotechnology Co., Ltd. (Beijing, PR China). Assay kits for reduced glutathione (GSH) and malondialdehyde (MDA)—a key marker of lipid peroxidation (LPO)—were acquired from Nanjing Jiancheng Bioengineering Institute. The compounds 2′,7-dichlorofluorescein diacetate (DCFH-DA), calcein AM, propidium iodide (PI), and annexin V-FITC were sourced from Solarbio Life Sciences. Antibodies employed in the experiments were supplied by YamayBio LLC and Beyotime Biotech, Inc. Analytical-grade chemicals and reagents were all used directly, with no further purification conducted. Ultrapure water (18.2 MΩ cm resistivity) served as the water source for all experiments.

Synthesis of ZB

2.2

ZB was synthesized using a one-pot synthesis method. First, 2.5 mg of PVP and 9.9 mg of Zn(NO_3_)2·6H_2_O were dissolved in 10 mL DMF and stirred for 30 min. Then, 2-nIm (4.7 mg), cnIm (3.9 mg), and BCH (1 mg) were added to the above-mentioned mixture and stirred for 30 min. The mixture was added to a tightly capped reactor and heated in an oven at 150 °C for 4 h. After cooling to room temperature, sequential washing of the product with DMF and ethanol was performed three times, with subsequent dispersion in ethanol as the final step. In addition, ZIF-82 was synthesized using a method similar to that of ZB, but without the addition of BCH.

Synthesis of ZB@HM

2.3

The hybrid membrane coating was performed using a co-extrusion method. A mini-extruder (Avestin LF-1, Canada) and a 400 nm-pore Nuclepore Track-Etch Membrane (Whatman) were employed to extrude the cancer cell membrane fraction, the bacterial membrane fraction, and the designed NPs 10 times. ZB@HM were collected and lyophilized for further experiments. Western Blotting and TEM were employed to confirm the successful HM package.

ZB@HM degradation studies

2.4

To simulate the acidic TME, ZB@HM was incubated in 1 mL of pH 6.5 PBS buffer across different time periods. At specified time points, the solutions were characterized via TEM to examine morphological properties.

Cellular internalization and cytotoxicity profile of ZB@HM

2.5

In vitro evaluation assay of cellular uptake. 96-well plates were inoculated with GL261 cells at a seeding density of 5000 cells per well, and the cells were cultured for 12 h under 37 °C and 5% CO_2_. Then they were incubated with Rhodamine B-conjugated ZB or ZB@HM (100 μL, equivalent ZIF-82 NPs dosage: 100 μg/mL). After a certain period of incubation (1, 2 and 4 h), the cells were thoroughly rinsed with pre-cooled PBS and stained with DAPI (1 μg/mL) for 15 min. Cellular internalization and distribution of Rhodamine B-labeled nanoparticles were assessed by fluorescence microscope. In vitro BBB penetration. A total of 2 × 10^5^ bEnd.3 cells were seeded into each insert placed in the upper chambers (pore size: 0.4 μm). The culture medium was refreshed every 2 days for 5–9 days, until the transmembrane resistance (measured using a Millicell ERS Voltohmmeter) exceeded 180 Ω/cm^2^, confirming the successful establishment of the in vitro simulated BBB model. Subsequently, GL261 cells were plated in the lower chamber and co-incubation for 24 h. Thereafter, 100 μL of fluorescein-labeled ZB or ZB@HM (with an equivalent ZIF-82 NPs dosage of 100 μg/mL) was added to the apical chamber, followed by a 6-h incubation period. Following this, GL261 cells in the bottom well were stained with DAPI (1 mg/mL) for 10 min, and fluorescein-labeled NPs in the bottom well were visualized using a fluorescence microscope to evaluate the BBB penetration capacity.

In vitro clonogenic testing

2.6

GL261 cells were plated in 6-well culture plates at a density of 400 cells per well. Following a 24-h incubation period, the cells were separately exposed to various nanodrugs for an additional 24 h. Subsequently, the aforementioned cells were incubated for 10 days in fresh culture medium. Finally, the cells were fixed and stained with crystal violet.

Detection of intracellular Zn2+in vitro

2.7

96-well plates were inoculated with GL261 cells at a concentration of 5000 cells per well. The plates were then placed in an incubator set to 37 °C and 5% CO_2_ for 12 h. Following initial preparation, the cells were incubated with distinct nanodrugs for 24 h. Thereafter, the cells were subjected to gentle rinsing three times using PBS. To evaluate intracellular Zn^2+^ concentrations, GL261 cells were stained with zinquin ethyl ester, washed with pre-cooled PBS, and visualized under a fluorescence microscope.

Assessment of intracellular ROS levels in vitro

2.8

A density of 5000 GL261 cells per well was used to inoculate 96-well plates, which were then cultured for 12 h under 37 °C and 5% CO_2_. Following initial preparation, the cells were incubated with different nanodrug samples for 24 h. Following this step, gentle washing with PBS was performed three times to get rid of unbound nanodrugs. To measure intracellular ROS in GL261 cells, the cells were stained with DCFH-DA, washed with pre-cooled PBS, and imaged under a fluorescence microscope.

In vitro mitochondrial membrane potential measurement

2.9

A density of 5000 GL261 cells per well was used for plating in 96-well plates, with subsequent culture at 37 °C and 5% CO_2_ for a 12-h period. Following this step, the cells were incubated alongside different nanodrugs for 24 h. These materials were then gently washed with PBS for three times. To measure mitochondrial membrane potential, GL261 cells were stained with JC-1, washed with pre-cooled PBS, and subsequently imaged using a fluorescence microscope.

Assay for intracellular LPO

2.10

A density of 1 × 10^5^ GL261 cells per well was used for seeding in 6-well plates and incubated overnight. Subsequently, the cells were incubated with various nanodrugs for a 24-h period. Following this incubation, the cells were gently rinsed three times with PBS. To assess intracellular LPO, GL261 cells were stained with the fluorescent probe C11-BODIPY^581/591^, followed by washing with pre-cooled PBS and imaging via fluorescence microscope.

Western Blotting

2.11

GL261 cells were exposed to distinct nanodrugs for 24 h and then washed with pre-cooled PBS. Then RIPA lysis buffer was employed for protein harvesting and BCA protein assay kit (Beyotime) was used for quantification. Protein samples were diluted to the same concentration and transferred to PVDF membranes by SDS-PAGE gradient gel separation. The membranes were incubated with antibodies against SLC3A2 (Beyotime), SLC7A5 (Beyotime), SLC7A11 (Beyotime), β-Actin (Beyotime), and GPX4 (Beyotime) at 4 °C, and then were HRP-conjugated goat anti-mouse/rabbit immune Globulin G (IgG) (Beyotime). Protein bands were taken using an enhanced chemiluminescence solution reaction.

ICD analysis

2.12

The expression of HMGB1 in the nucleus of GL261 cells was measured by immunofluorescence. GL261 cells after dealt with different nanodrugs for 24 h were fixed and permeabilized. Then, these cells were co-incubated with anti-HMGB1 antibody (Beyotime) overnight, followed by incubation with FITC-conjugated Goat anti-Rabbit reagents (Beyotime) for 30 min. GL261 cells were then co-stained with DAPI (Servicebio) and visualized by fluorescence microscope. The expression of CRT on the membrane of GL261 cells was measured by immunofluorescence. GL261 cells after dealt with different nanodrugs for 24 h were fixed. Then, these cells were co-incubated with anti-CRT antibody (Beyotime) overnight, followed by incubation with FITC-conjugated Goat anti-Rabbit antibodies (Beyotime) for 30 min. GL261 cells were then co-stained with DAPI (Servicebio) and visualized by fluorescence microscope. ATP assay kit (Beyotime) was used to quantify extracellular ATP secretion.

Assessment of biodistribution and antitumor activity of ZB@HM using an orthotopic glioma model

2.13

Biodistribution analysis. Tumor-bearing mice were euthanized via cervical dislocation at 12, 24, and 48 h post-intravenous injection of ZB@HM. Thereafter, the hearts, livers, spleens, lungs, kidneys, and brains of the mice were dissected and weighed. Aqua regia (HNO_3_:HCl = 1:3) was added for digestion over 7 days. ICP-MS was employed to assess Zn content. Animal model of tumors. For the construction of the glioma mouse model, female C57BL/6 mice (6–8 weeks old, 15–20 g) were utilized. Anaesthetized mice were fixed at a stereotaxic apparatus and then the head was sterilized with 75 % ethanol. A burr hole was drilled 2 mm right of the bregma and 1 mm rostral to the coronal suture. Then GL261 cells (1 × 10^6^ in 7 μL PBS) were injected 3 mm deep into the brain tissue. The protocol for animal experiments was evaluated and authorized by the Experimental Animal Ethics Committee of The First Hospital of Jilin University, with the ethical certification number 20240227-55. All experiments were performed in compliance with the pre-approved guidelines. Anti-tumor in vivo. To evaluate the therapeutic efficacy of different nanodrugs in orthotopic glioma, we established mice orthotopic glioma model and started treatment at 10 days after inoculating GL261 cells into mice brain. Intravenous injection was used to administer nanodrugs to mice at a dosage of 20 mg/kg, and T2W-MRI was employed to measure the tumor volume of each group across the full duration of the treatment. Body weight measurements of the mice were conducted every 48 h. On day 12, tumor-bearing mice were sacrificed via cervical dislocation, and their hearts, livers, spleens, lungs, kidneys, and brains were dissected. These harvested organs were immediately fixed in formalin and then paraffin-embedded for H&E staining procedures.

In vivo assessments of macrophage polarization

2.14

To assess the macrophage polarization in vivo, we obtained the glioma tissues after different treatments and made into the single-cell suspension. Cell suspensions were stained with FITC-conjugated anti-F4/80, PE-conjugated anti-CD86, PerCP/Cyanine5.5-conjugated anti-CD206, and APC-conjugated anti-CD11b antibodies for a 30-min incubation period. Thereafter, the cells that had been stained were harvested and subjected to flow cytometric examination.

DC maturation in vivo

2.15

These antibodies were purchased from BioLegend unless otherwise mentioned. The maturation of DCs in lymph nodes were measured. Initially, the target tissue was processed into a cell suspension and incubated with anti-CD80-FITC, anti-CD86-PE, and APC-conjugated anti-CD11b antibody (Elabscience Biotechnology Co., Ltd.) at 4 °C for 30 min. Unbound antibodies were eliminated via centrifugation. Following this, the cells were washed with PBS, resuspended in fresh PBS, and subjected to flow cytometric analysis.

In vivo assessments of T cell subsets

2.16

To assess CD8^+^ T cells and CD4^+^ T cells within lymph node and tumor tissues, the latter were processed into cell suspensions and stained with anti-CD3-FITC, anti-CD4-APC, and anti-CD8a-PE for 30 min. Then, cells were centrifuged to remove excess antibodies and washed several times with PBS before being redispersed. Prepared cells were examined by flow cytometry.

Statistical analysis

2.17

The option of sample size was determined based on the preliminary data of pilot experiments as well as results from previous literature. All animal experiments were conducted after randomization. Data are presented as the mean ± standard deviation (SD) in accordance with the specified requirements. A one-way ANOVA was adopted for the analysis to conduct statistical analysis of the data, while a two-tailed Student's t-test was utilized to assess the significance of differences between the two groups. Statistical analyses were performed using Origin 2023 software, and asterisks are used to denote significant differences (∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, and ∗∗∗∗P < 0.0001).

Results and discussion

3

Fabrication and characterization of ZB@HM

3.1

ZIF-82 nanoparticles (NPs) were synthesized via a literature-reported procedure, and transmission electron microscopy (TEM) and scanning electron microscope (SEM) characterization demonstrated that the NPs possessed a spherical structure with an average size of 75 nm (Fig. 1a and S1) [11]. The X-ray diffraction (XRD) patterns showed the diffraction peaks of ZIF-82 NPs matched well with the simulation data (Fig. S2). BCH-loaded amorphous ZIF-82 NPs, denoted as ZB NPs, were synthesized via a one-pot approach and exhibited excellent monodispersity with a relatively uniform size distribution-both critical prerequisites for the subsequent biomimetic coating (Fig. 1b). The XRD patterns implied the amorphous structures of ZB NPs (Fig. S2). Nitrogen adsorption-desorption isotherm analyses demonstrated that the ZB NPs exhibited a specific surface area of 231.531 m^2^g^-1^ and an average pore diameter of 3.407 nm (Fig. 1f). Two distinct peaks were observed in the Zn 2p spectrum at around 1022 eV and 1045 eV, corresponding to Zn ^2^p_3/2_ and Zn ^2^p_1/2_, respectively (Fig. 1g). All the above results indicated the successful synthesis of ZB NPs. To improve biosafety and bio-barrier penetration, ZB NPs were coated with a hybrid membrane, derived from E. coli bacterial membranes and glioma tumor cell membranes, via a co-extrusion method to obtain the ZB@HM nanotuner [29]. Bacterial membrane enhances BBB traversal, while tumor cell membrane confers homing capability for targeted accumulation [30]. As shown in Fig. 1c, a distinct hybrid membrane structure is observable on the surface of ZB@HM nanotuner with a diameter of less than 90 nm. The presence of the hybrid membrane structure is expected to endow ZB@HM nanotuner with unique properties such as improved biocompatibility and enhanced cellular uptake. Moreover, sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was used to ascertain the presence of E. coli- and GL261-derived hybrid membrane proteins on the surface of ZB. Fig. 1d demonstrates that there exist no notable variations in protein bands between the hybrid membrane and ZB@HM, indicating the successful coating of the membrane on the surface of ZB. High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) was employed to provide additional verification that the hybrid membrane coating process had no adverse effects on the structural integrity of ZB NPs, using ZB@HM as the target sample. As shown in Fig. 1e, it revealed the homogenous distribution of Zn, which provided strong evidence for a well-controlled synthesis process of the ZB@HM. After the hybrid membrane coating process was implemented, the Zeta potential exhibited a notable shift, decreasing from −3.1 mV to −8.8 mV, whereas the average hydrodynamic diameter rose from 98.7 nm to 122.1 nm (Fig. 1h and i). The successful formation of ZB@HM was further verified through these experimental results. Quantification experiments were conducted to evaluate the drug loading capacity and encapsulation efficiency of ZB@HM for BCH, yielding values of 11.7% and 56.5%, respectively (Table S1). This result confirms that ZB@HM serves as a viable candidate for drug delivery applications. ZB@HM degrades into Zn^2+^, 2-nim, and BCH. The degradation mechanism was investigated via TEM characterization. Notably, marked structural disintegration was detected when ZB@HM was exposed to pH 6.5 for 30 min, and ZB@HM accomplished the entire degradation process within an 8-h duration (Fig. 1j). Over a 7-day incubation period, the mean diameter of ZB@HM showed no statistically significant changes when suspended in deionized water (DI water), phosphate-buffered saline (PBS), or Dulbecco's modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (Fig. 1n). These observations indicate that ZB@HM maintains favorable stability under physiological conditions. To evaluate the component release performance of ZB@HM, we analyzed the release behaviors of Zn^2+^, BCH, and 2-nim in PBS under three varying pH conditions: 5.0, 6.5, and 7.4. When ZB@HM was dispersed in acidic PBS (pH 5.0 and 6.5) for 24 h, the cumulative release of Zn^2+^, BCH, and 2-nitroimidazole from ZB@HM exceeded 80% (Fig. 1k, l and m). On the other hand, when quantified under physiological conditions (pH 7.4), the release percentages of Zn^2+^, BCH, and 2-nim were limited to no more than 20%. This pH-dependent release mechanism promotes drug release that is specific to tumor tissues, thus lessening the occurrence of non-targeted side effects. Next, the depletion of NADPH by 2-nim was further verified by detecting NADPH's characteristic absorbance peak at 340 nm. As presented in Fig. 1o and p and S3, a significant reduction in NADPH absorbance was noted following the addition of ZB@HM, demonstrating its robust NADPH-consuming capacity in a time-dependent manner. Moreover, as the reaction time of ZB@HM increased, the decrease in NADPH absorbance accelerated, which further confirms the prominent NADPH-consuming activity of ZB@HM (Fig. 1q and r).Fig. 1a) TEM image of ZIF-82, scale bar: 500 nm; TEM images of b) ZB and c) ZB@HM, scale bar: 100 nm; d) SDS-PAGE analysis for membrane proteins; e) HAADF-STEM image and elemental mapping of Zn in ZB@HM; f) N_2_ adsorption-desorption isotherms of ZB; g) Zn 2p spectrum of ZB; h) Zeta potential and i) hydrodynamic diameters of ZB and ZB@HM. Data are mean ± SD; j) TEM images of ZB@HM in a pH 6.5 buffer solution at 0 min, 30 min, 1 h, and 8 h; The time-dependent release percentage of k) BCH, l) Zn^2+^ and m) 2-nim from ZB@HM at different pH values (pH 7.4, pH 6.5, pH 5.0); n) Hydrodynamic diameters of ZB@HM measured in different media (DI Water, PBS, DMEM (10% FBS)) for 7 days; o) Absorbance of NADPH following treatment with different concentrations of ZB@HM (pH 6.5); p) Relative content of NADPH following treatment with different concentrations of ZB@HM (pH 6.5); q) Absorbance of NADPH following treatment with ZB@HM (pH 6.5) at different reaction time; r) Relative content of NADPH following treatment with ZB@HM (pH 6.5) at different reaction time.Fig. 1

In vitro BBB-permeability and antitumor efficacy

3.2

Biocompatibility is a critical consideration for the clinical translation of NPs. The cytotoxicity of ZB@HM was first examined through the counting kit-8 (CCK-8) assay. Following a 24-h incubation period with high-concentration ZB@HM under physiological conditions, the viability of the treated cells was maintained at over 95% (Fig. 2d). This observation confirms that ZB@HM possesses favorable biosafety characteristics. Functioning as a critical physiological barrier, the BBB imposes limitations on the delivery of the majority of substances to the cerebral. Thus, crossing the BBB constitutes a fundamental prerequisite for the effective performance of ZB@HM. To establish the BBB model, a transwell system was adopted, where the upper chamber was dedicated to the cultivation of a confluent bEnd.3 cell monolayer, and the lower chamber was reserved for the seeding of GL261 cells (Fig. 2a). The integrity of the BBB model was validated via the measurement of transepithelial/transendothelial electrical resistance (TEER). Furthermore, TEER values remained stable following co-cultivation of the BBB model with ZB or ZB@HM, which indicates that the BBB model was not disrupted (Fig. S4). As depicted in Fig. 2b, after ZB was added to the upper chamber and subjected to 6 h of incubation, GL261 cells in the lower pores displayed only a slight fluorescent signal. Conversely, treatment with ZB@HM led to a substantial enhancement in fluorescent intensity, thereby confirming that ZB@HM possesses strengthened BBB penetration capability. Subsequently, the uptake capacity of GL261 cells for Rhodamine B-labeled ZB or ZB@HM was evaluated at different time points using a fluorescence microscope. The intracellular uptake of ZB@HM was significantly higher than that of ZB, which supports the conclusion that hybrid membrane modification boosts the capacity of cells to phagocytose nanoparticles (Fig. 2c).Fig. 2a) Schematic diagram of in vitro BBB model for assessing the capacity of NPs to across the BBB; b) Fluorescent microscopy images showing localization of FITC-labeled NPs in GL261 cells following transport across an in vitro BBB model. Scale bar: 50 μm; c) Fluorescent microscopy images showing cellular uptake of Rhodamine B-labeled NPs by GL261 cells assessed at the designated time points. Scale bar: 50 μm; d) Viability of BV2 and GL261 cells after exposure to NPs of different concentrations in environments with distinct pH values; e) Colony formation assay images of GL261 cells across distinct groups; f) Quantification of colony numbers in the colony formation assay for different treatment groups (1: Control, 2: BCH, 3: ZIF-82, 4: ZB, 5: ZB@HM); g) Microscopic fluorescence images of GL261 cells in different groups, which were stained using Calcein-AM and PI. Scale bar: 50 μm; h) Apoptosis in GL261 cells was analyzed via flow cytometry following staining with Annexin V-FITC and PI. All data are presented as mean ± SD, with statistical significance defined as ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001.Fig. 2

While the acidic medium alone (pH 6.5, Con group) showed negligible cytotoxicity, ZB@HM exhibited a significant inhibitory effect on the cellular viability of GL261 cells under this TME condition. Specifically, at a concentration of 200 μg/mL, the viability of GL261 cells was reduced to less than 20% (Fig. 2d). In the colony formation assay, ZB@HM also showed a more prominent ability to suppress colony formation compared with other treatment groups (Fig. 2e and f). Cell staining by calcein AM and propidium iodide (PI) provided a more intuitive verification of anticancer effect. Overall, ZB@HM demonstrated potent antitumor activity (Fig. 2g). Apoptosis, a programmed cell death mechanism crucial for tumor suppression, was evaluated via Annexin V-FITC/PI staining. The results demonstrate that ZB@HM triggered the most potent pro-apoptotic effect among all treatments, inducing the highest levels of both early and late apoptosis in GL261 cells (Fig. 2h). Collectively, these findings indicate that the ZB@HM is a promising nanomaterial with significant BBB permeability, glioma-targeting specificity, and potent cytotoxicity specifically triggered by the acidic TME.

Activating ferroptosis via key substrate consumption

3.3

We further pursued an in-depth exploration of the intrinsic mechanisms that mediate the cytotoxicity of ZB@HM against glioma. Zinc ions are known to induce mitochondrial impairment, a process that in turn suppresses adenosine triphosphate (ATP) synthesis and elicits the production of reactive oxygen species (ROS) [31,32]. The burst release of Zn^2+^ in the acidic tumor microenvironment is a primary driver of cytotoxicity, while the material remains biologically inert in normal physiological conditions. To assess Zn^2+^ release from various NPs in GL261 cells, we used zinquin ethyl ester. As shown in Fig. 3a, blue fluorescence (indicating Zn^2+^) was observed in the ZIF-82, ZB, and ZB@HM groups, unlike the control. These results confirm Zn^2+^ release from these NPs, implicating a potential mechanism for their impact on mitochondrial dysfunction and oxidative stress. Mitochondrial damage was assessed using the fluorescent probe JC-1. As shown in Fig. 3b, the control group exhibited strong red fluorescence, indicating intact mitochondrial membrane potential. In contrast, ZB@HM showed the highest green fluorescence intensity than other treatment groups, suggesting the most severe disruption of mitochondrial function. This enhanced effect is likely attributable to more efficient cellular uptake. Subsequently, the ROS generation capability of various nanomaterials was evaluated using DCFH-DA. As presented in Fig. 3c, negligible fluorescence signals were detected exclusively in the control group. In contrast, distinct green fluorescence was observed in GL261 cells treated with ZB@HM under acidic conditions (pH = 6.5), indicating elevated ROS production within the TME-mimetic setting. This ROS burst is critical for inducing oxidative stress and subsequent cell death.Fig. 3a) Fluorescent microscopy images of GL261 cells in different groups stained with Zn^2+^ indicator probe, coupled with the quantification of the probe's fluorescence intensity (1: Control, 2: BCH, 3: ZIF-82, 4: ZB, 5: ZB@HM). Scale bar: 50 μm; b) Fluorescent microscopy images of GL261 cells in different groups stained with JC-1, and the quantification of the ratio between JC-1 aggregates and monomers. Scale bar: 50 μm; c) Fluorescent microscopy images of GL261 cells in different groups stained with DCFH-DA, and quantification of the fluorescence intensity of DCFH-DA. Scale bar: 50 μm. Bar graph showing d) the ratio of NADP^+^/NADPH, e) the relative cysteine content, f) the relative methionine content, g) the relative GSH content, h) the relative total thiol content, i) the relative Trx content, j) the relative GPX4 activity, and k) the relative MDA content in GL261 cells for different treatment groups, 1: Control, 2: BCH, 3: ZIF-82, 4: ZB, 5: ZB@HM and 6: ZB@HM in hypoxia microenvironment. l) Western blot analysis of proteins in vitro related to ferroptosis, including SLC3A2 (4F2hc), SLC7A11 (xCT), SLC7A5 (LAT1), GPX4, and β-Actin; m) Fluorescence microscopic images of GL261 cells across various groups stained with C11-BODIPY. 1: Control, 2: BCH, 3: ZIF-82, 4: ZB, and 5: ZB@HM. Data are mean ± SD. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001.Fig. 3

To assess the effectiveness of the ZB@HM in depleting NADPH, the NADP^+^/NADPH ratio was measured. Under hypoxic conditions, ZB@HM treatment induced a substantial increase in the relative NADP^+^/NADPH ratio, which reached a value of 2.45, indicating effective NADPH consumption, which consequently disrupted reduced equivalent of tumor cells and the GSH regeneration pathway (Fig. 3d). As shown in Fig. 3e, ZB@HM treatment induced a significant depletion of cysteine, which was more pronounced under hypoxic conditions. The expression levels of SLC7A11 and SLC3A2, the main amino-acid transporters responsible for cysteine transport, showed a downward trend (Fig. 3l) [33]. This process leads to the depletion of intracellular cysteine, thus interfering with the direct synthesis pathway of GSH. Furthermore, in response to the TME, ZB@HM released BCH, which effectively suppressed the indirect GSH synthesis pathway. This was corroborated by a significant decrease in methionine levels across treatment groups, with the most pronounced reduction observed in ZB@HM-treated cells under hypoxia (Fig. 3f). Consistent with this, expression of methionine transporters SLC7A5 and SLC3A2 was also reduced (Fig. 3l). These results collectively demonstrate that ZB@HM efficiently blocks direct, indirect, and regeneration pathway of GSH synthesis through multi-regulation of NADPH, cysteine and methionine metabolism. GSH serves as a critical antioxidant in gliomas. Depletion of GSH can enhance cellular susceptibility to oxidative stress, which may underlie its therapeutic potential. A marked downregulation of GSH levels induced by ZB@HM under hypoxia was further validated using a commercial assay kit (Fig. 3g). Moreover, ZB@HM significantly depleted the total thiol, particularly under hypoxic microenvironments, reducing the relative content to approximately 50% (Fig. 3h). ZB@HM had the ability to substantially deplete reduced Trx (Fig. 3i). The depletion of reduced Trx further disrupts the cellular redox balance, making cells more vulnerable to oxidative damage.

Subsequently, we investigated the mechanisms through which ZB@HM induces ferroptosis. The activity and expression of GPX4 showed a decreasing trend (Fig. 3j and l). Subsequently, we measured LPO levels in GL261 cells treated with various NPs. Significantly brighter green fluorescence was observed in the ZB@HM group compared to other groups (Fig. 3m), indicating enhanced LPO accumulation. This effect can be attributed to the suppression of multiple GSH synthesis pathways due to ZB@HM-induced depletion of methionine, cysteine, and NADPH. Furthermore, changes in malondialdehyde (MDA), a major LPO end product, were consistent with the LPO findings (Fig. 3k). Collectively, the increased Fe^2+^ levels, GSH depletion, reduced GPX4 activity, and subsequent LPO accumulation strongly indicate ferroptosis induction by ZB@HM. As a non-apoptotic form of cell death, ferroptosis represents an emerging therapeutic target in oncology. These results demonstrate the potential of ZB@HM as a ferroptosis-inducing agent for glioma treatments.

Initiation of immunogenic cell death and its immunomodulatory consequences

3.4

ICD is intricately intertwined with the induction of antitumor immunity, serving as a pivotal and indispensable element in this process. Ferroptosis holds significant potential in initiating an adaptive immune response via the induction of ICD. A key molecular characteristic of ICD is the release of DAMPs, typically manifested as the extracellular release of high mobility group box 1 (HMGB1), the cell surface exposure of calreticulin (CRT), and the liberation of extracellular adenosine triphosphate (eATP). To explore this, we investigated the release of DAMPs from GL261 glioma cells under different treatment conditions. As shown in Fig. 4a and b, a notably more efficient HMGB1 release was observed in the ZB@HM group compared to other treatments, evidenced by its significantly reduced intracellular level. Compared to the control, other treatments of GL261 cells induced CRT exposure, with the most prominent effect observed in the ZB@HM group (Fig. 4c and d). Furthermore, all treatment groups significantly enhanced ATP secretion, with ZB@HM again exhibiting the highest level (Fig. 4e). These findings indicate that ferroptosis-induced ICD can be modulated by different treatments. Specifically, ZB@HM was proved to be the most effective ferroptosis inducer, potently stimulating the release of DAMPs, including HMGB1, CRT exposure, and ATP secretion, thereby eliciting the most robust ICD response.Fig. 4. Fluorescent microscopy images of GL261 cells in different groups stained with a) HMGB1, c) CRT. Scale bar: 50 μm; Bar graph depicting the mean fluorescence intensity (MFI) of b) HMGB1, and d) CRT in different treatment groups (1: Control, 2: BCH, 3: ZIF-82, 4: ZB, 5: ZB@HM); e) The relative extracellular ATP (eATP) content in GL261 cells for different treatment groups; f) The relative SAM content in GL261 cells for different treatment groups; g) Western blot analysis of proteins in vitro related to immune, including PD-L1, GAPDH, H3K4Me3, and Histone H3; h) Representative flow cytometry analysis of CD86 and F4/80 expression in mice with different groups, and bar graph showing the percentage of M_1_ in GL261 cells for different treatment groups; i) Representative flow cytometry analysis of CD206 and F4/80 expression in mice with different groups, and bar graph showing the percentage of M_2_ within GL261 cells among various treatment groups. 1: Control, 2: BCH, 3: ZIF-82, 4: ZB, and 5: ZB@HM. Data are mean ± SD. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001.Fig. 4

To validate the 'methionine redistribution' effect, we utilized a Transwell co-culture system simulating the nutrient-competitive TME (Fig. S5). Unlike free BCH, which exacerbated T cell death, ZB@HM significantly reduced T cell apoptosis while suppressing tumor survival. This confirms that ZB@HM effectively relieves metabolic pressure on T cells by selectively restricting tumor methionine uptake. Reduced methionine availability markedly lowers SAM levels in tumor cells (Fig. 4f). As the primary methyl donor, SAM directly regulates histone H3 lysine 4 trimethylation (H3K4Me3), an epigenetic mark controlling immune-related gene expression. Consequently, this reduction leads to decreased H3K4Me3 methylation and downregulates PD-L1 expression (Fig. 4g). By attenuating PD-L1-mediated immunosuppression, this metabolic-epigenetic axis enhances anti-tumor immunity. These results reveal a core mechanism linking methionine metabolism to epigenetic reprogramming and immune evasion, providing a rationale for multi-modal targeting in cancer immunotherapy. Flow cytometry analyses were employed to quantify both the proinflammatory M_1_-phenotype macrophages and the protumor M_2_-phenotype macrophages [34]. In the treatment groups, the proportions of M_1_-phenotype macrophages were consistently higher compared to the control group, while the proportions of M_2_-phenotype macrophages exhibited an opposite trend (Fig. 4h and i). In the ZB@HM group, the M_1_/M_2_ ratio reached up to 4, surpassing that of other groups (Fig. S6). This phenomenon can be ascribed to the enhanced ICD and the presence of tumor-associated antigens in the ZB@HM-treated group. The increased ICD in the ZB@HM group Triggers the release of a greater number of DAMPs as well as tumor-associated antigens which can polarize macrophages towards the M_1_ phenotype. The M_1_-polarized macrophages are more effective in killing tumor cells and promoting an antitumor immune response, while the decrease in M_2_-phenotype macrophages reduces the immunosuppressive TME. This shift in the M_1_/M_2_ ratio in the ZB@HM group is a key finding, as it indicates that the ZB@HM treatment not only induces ICD but also has a significant impact on modulating the immune cell composition in the TME, potentially enhancing the overall antitumor immune response.

Transcriptome analysis of ZB@HM

3.5

Given the potent glioma cell-killing activity of ZB@HM, we employed RNA sequencing (RNA-seq) to dissect the underlying molecular mechanisms. Transcriptomic profiling identified 7494 differentially expressed genes (DEGs) in ZB@HM-treated cells, with 5305 upregulated and 2189 downregulated (|log2(FoldChange)| > 1, padj <0.05; Fig. 5a). Gene Ontology (GO) enrichment analysis of these DEGs revealed 15 significantly perturbed pathways, prominently including the regulation of GSH metabolic processes (GO:0006749, 0006575, 0006790, 0004364, 0004602, 0043295) and immune response activation (GO:1903131, 0030098, 0042613, 0042611, 0042101, 0001772; Fig. 5b). These transcriptional changes were consistent with our functional observations, reflecting depletion of precursor pools in GSH biosynthetic pathways and intracellular redox imbalance—key events driving tumor cell ferroptosis. Additionally, the enrichment of immune response pathways prompted further investigation into downstream immunomodulatory effects.Fig. 5a) The volcano map. b) Gene Ontology enrichment analysis about representative pathways among the identified gene set. c) KEGG analysis about representative pathways among the identified gene set. d) GSH metabolism. e) The PD-1 and PD-L1 checkpoint pathway in tumor.Fig. 5

Analysis based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) further confirmed ZB@HM-induced activation of critical biological processes, including Th1/Th2 cell differentiation, neutrophil extracellular trap formation, GSH metabolism, and cytokine-cytokine receptor signaling (Fig. 5c). Collectively, these data indicate that ZB@HM disrupts multiple ferroptosis defense systems while triggering immunomodulatory responses. Transcriptome analysis highlighted a striking upregulation of GSH metabolism-related genetic signatures (Fig. 5d). Concurrently, ZB@HM treatment reshaped transcriptional programs within the TIME. Notably, PD-L1, a critical mediator that drives CD8^+^ T-cell suppression in the TME, showed diminished expression in glioma cells—accompanied by increased levels of pivotal immunomodulatory genes including TNF-α, IFN-γ, and IL-6 (Fig. 5e). These findings directed our focus toward mechanisms underlying ZB@HM-mediated TIME remodeling and its downstream immune consequences.

In vivo antitumor efficacy of ZB@HM

3.6

Given the encouraging in vitro therapeutic efficacy, we evaluated the in vivo antitumor effectiveness of different treatment modalities. The biodistribution of ZB@HM was evaluated via inductively coupled plasma mass spectrometry in this study. Significant accumulation of ZB@HM was observed in the brain, demonstrating its effective BBB penetration and providing a foundation for subsequent in vivo glioma therapy (Fig. S7). To assess the therapeutic effect of ZB@HM, orthotopic glioma models were employed. Glioma-bearing mice were randomly assigned to five groups: (1) Control group, (2) BCH group, (3) ZIF-82 group, (4) ZB group, and (5) ZB@HM group. The treatment protocol is illustrated in Fig. 6a, detailing the time points of drug administration and tumor size measurement via MRI. In the control group, the tumor volume exhibited a significant increase over time. In the BCH and ZIF-82 groups, minimal inhibition of orthotopic gliomas was observed, suggesting that BCH or ZIF-82 alone failed to exert a therapeutic impact. The lack of significant tumor inhibition in the BCH and ZIF-82 single-agent groups underscore the importance of the combination strategy in ZB@HM. In contrast, tumor growth was inhibited in the ZB@HM group demonstrating the most pronounced inhibitory effect, presenting the smallest tumor volume after treatment. The synergistic effect between the components of ZB@HM is likely responsible for its enhanced antitumor activity. Compared with ZB alone, the potent antitumor efficacy of ZB@HM is largely attributable to its hybrid membrane coating, which enhances BBB penetration and facilitates targeted delivery to orthotopic glioma sites (Fig. 6b and e). This result was consistent with the observations from brain tissue sections of mice with glioma xenografts (Fig. 6c). After the ZB@HM treatment, distinct cellular damage and sparser glioma tissue were observed in GL261 cells by hematoxylin and eosin (H&E) staining (Fig. 6d). Western blotting analysis demonstrated a significant reduction in the expression of SLC3A2, SLC7A5, SLC7A11, and GPX4 in tumor tissues of the ZB@HM group (Fig. 6f). These findings strongly suggest that ZB@HM can trigger ferroptosis in vivo. The observed activation of ferroptosis in the ZB@HM-treated group, as evidenced by the downregulation of key proteins related to ferroptosis, provides mechanistic insights into its antitumor action.Fig. 6a) Timeline illustration depicting the experimental procedure for glioma-related research in mice; b) MRI scans of mice brains in different treatment groups at various time points; c) Representative micrographs of brain tissue sections derived from distinct treatment groups. scale bar = 1000 μm; d) High-magnification histological micrographs of brain tissues derived from distinct treatment groups. scale bar = 100 μm; e) Relative tumor volume curves of brain tumor from different treatment groups. f) Western blot analysis of proteins in vivo related to ferroptosis, including SLC3A2 (4F2hc), SLC7A11 (xCT), SLC7A5 (LAT1), GPX4, and β-Actin. g) Body weight of mice.Fig. 6

In pre-clinical studies, maintaining the balance between therapeutic efficacy and safety is of utmost importance. Notably, throughout the entire treatment period, no statistically significant alterations were observed in the body weight of the experimental mice, and no visible damage was observed in the tissue sections of major organs. This indicates the low systemic toxicity of ZB@HM to mice (Fig. 6g and S8). The hemolysis test results further validated the excellent hemocompatibility of ZB@HM in blood circulation (Fig. S9). Additionally, blood routine and blood biochemical analyses showed values within the normal range, further confirming the negligible in vivo impact of ZB@HM (Figs. S10 and S11). The low systemic toxicity and good hemocompatibility of ZB@HM are crucial for its translational potential. The fact that ZB@HM shows no significant adverse effects on body weight, major organs, blood cells, and blood biochemistry indicates that it has the potential to be developed into a safe and effective therapeutic agent for gliomas.

In vivo antitumor immune mechanisms of ZB@HM

3.7

We establish a link between the tumor immune microenvironment (TIME) and methionine availability. Via extensive database analysis, high SLC3A2, SLC7A5, SLC7A11, and GPX4 levels were negatively correlated with an active TIME—characterized by increased infiltration of dendritic cells (DCs), CD4^+^ T cells, CD8^+^ T cells, and M1 macrophages, alongside decreased M2 macrophages (Fig. 7a and S12). Understanding the in vivo immune response triggered by ICD is crucial for developing effective cancer therapies. To this end, we harvested lymph nodes, spleens, and tumors for subsequent flow cytometry analysis. DAMPs release is known to synergistically promote DCs maturation, which is a pivotal step in initiating antitumor immunity. The maturation rate of DCs in the tumor-draining lymph nodes (TDLNs) was examined and found to be 57.1% in the ZB@HM group, significantly exceeding that of the ZB group (43.7%), the ZIF-82 group (30.0%), the BCH group (27.5%), and the control group (21.7%) (Fig. 7b). This substantial increase in DC maturation within the ZB@HM group indicates that the ZB@HM has a potent effect on promoting this key immunological process. The initiation, regulation, and sustenance of immune responses rely critically on mature DCs. To evaluate T cell-mediated immunity after different treatments, we quantified the proportions of helper CD4^+^ T cells and cytotoxic CD8^+^ T cells in mice. Flow cytometry analysis of downstream T cells in tumors showed that CD4^+^ T cells and CD8^+^ T cells could be effectively activated by ZB@HM, which are crucial for inhibiting tumor growth (Fig. 7c and d). In addition, in the spleen of mice from the ZB@HM group, CD8^+^ T and CD4^+^ T cells accounted for approximately 24.9% and 36.0% respectively, representing the highest levels in all groups (Fig. 7e and f). Moreover, ZB@HM could further promote DC maturation and the polarization of tumor-associated macrophages (TAMs) through enhanced ferroptosis. These findings suggest that by potently inducing ferroptosis and downregulating methionine uptake in tumor cells, the resultant increase in methionine availability enhances CD8^+^ T cell function, thereby effectively activating ICD-induced antitumor immunity and alleviating immunosuppressive TME.Fig. 7a) Relationship of SLC7A5 expression and immune cell infiltration levels based on information extracted from the Tumor Immune Estimation Resource (TIMER) database. b) Representative assessment of CD86 and CD80 expression in mouse tumor tissues via flow cytometry with different groups, and bar graph showing the percentage of DC cells for different treatment groups (1: Control, 2: BCH, 3: ZIF-82, 4: ZB, 5: ZB@HM); c) Representative Flow cytometric analysis of CD3 and CD4 expression in mouse tumor tissue, and bar graph showing the percentage of CD4^+^ T cells (1: Control, 2: BCH, 3: ZIF-82, 4: ZB, 5: ZB@HM); d) Representative flow cytometry-based assessment of CD3 and CD8 expression patterns in murine tumor tissue, and bar graph showing the percentage of CD8^+^ T cells (1: Control, 2: BCH, 3: ZIF-82, 4: ZB, 5: ZB@HM); e) Representative flow cytometric analysis of CD3 and CD4 expression in the spleens of mice from distinct experimental groups, and bar graph showing the percentage of CD4^+^ T cells for different treatment groups (1: Control, 2: BCH, 3: ZIF-82, 4: ZB, 5: ZB@HM); f) Representative flow cytometric analysis of CD3 and CD8 expression in mouse spleens from various experimental groups, and bar graph showing the percentage of CD8^+^ T cells for different treatment groups (1: Control, 2: BCH, 3: ZIF-82, 4: ZB, 5: ZB@HM); (g-i) The release of cytokines IL-6, TNF-α, and IFN-γ measured by ELISA assay. Data are mean ± SD. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001; j) Western blot analysis of proteins in vivo related to immune responds, including PD-L1, H3K4Me3, GAPDH, and Histone H3.Fig. 7

To further validate our findings, we performed cytokine-related ELISA on serum samples. In agreement with flow cytometric data, quantitative analysis revealed that ZB@HM treatment induced the most robust cytokine response, with significantly increased production of interferon-γ (IFN-γ), interleukin-6 (IL-6), and tumor necrosis factor-α (TNF-α) compared to other treatment regimens (Fig. 7g–i). This indicates a more robust antitumor immune response in ZB@HM group. The elevated cytokine levels can have multiple effects, including attracting immune cells to the tumor site, enhancing the cytotoxic activity of T cells, and modulating the TME to a more immunostimulatory state. In line with the in vitro findings, A significant reduction in the expression of H3K4Me3 and PD-L1 was also observed in the ZB@HM group (Fig. 7j). In conclusion, our comprehensive results show that ZB@HM not only promotes DC maturation and T-cell activation but also modulates the cytokine profile in a way that supports a strong antitumor immune response. These results underscore the potential of ZB@HM as a viable therapeutic agent for glioma therapy, as it leverages the antitumor immune responses triggered by ICD.

Conclusions

4

We developed a biomimetic nanotuner based on sequentially gated amorphous MOF, achieving the cascade depletion two key substrates to disrupt ferroptosis defense systems and reprogramming of methionine for boosting ferroptosis-mediated immunotherapy of glioma. The E. coli/glioma hybrid cell membrane functionalization make ZB@HM with good ability of BBB traversal and selective accumulation in glioma regions. Under acidic TME, the amorphous MOF releases the electrophilic ligand, which is subsequently reduced by NADPH in hypoxic regions to its active form. The activated ligand covalently binds to thiol groups in cysteine, leading to coordinated depletion of both NADPH and cysteine. This dual depletion initiates potent ferroptosis and promotes ICD. In parallel, ZB@HM selectively inhibits methionine uptake in tumor cells, thereby reprogramming metabolic partitioning to favor methionine utilization by T cells. Consequently, this approach reduces intra-tumoral SAM levels, downregulates immune checkpoint expression, and alleviates T cell exhaustion, leading to enhanced antitumor immunity. Experimental data obtained from both in vitro and in vivo settings provide evidence that ZB@HM potently induces ferroptosis and remodels the immunosuppressive microenvironment of glioma. Together, these findings demonstrated ZB@HM a promising candidate for augmenting ferroptosis-driven immunotherapy in glioma.

CRediT authorship contribution statement

Bin Wang: Conceptualization, Investigation, Methodology, Validation, Visualization, Writing – original draft. Na Yin: Conceptualization, Supervision, Writing – review & editing. Xinrui Liu: Supervision, Writing – review & editing. Yi Guan: Methodology. Haiyang Xu: Methodology. Ying Yu: Methodology. Yang Bai: Visualization. Yue Cao: Investigation. Ziqian Wang: Investigation. Shiqi Bai: Investigation. Shaopeng Zhang: Investigation. Donghao Qu: Investigation. Wanying Li: Investigation. Zhijia Lv: Visualization. Yunqian Li: Supervision. Hongquan Yu: Funding acquisition, Supervision, Writing – review & editing. Yinghui Wang: Conceptualization, Funding acquisition, Supervision, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.